March 17, 20253 minutes
My notable projects during undegrad studies.
This website is currently under development. Tech to be tried: Next.js, React, Tailwind CSS, Vercel Hosting, MongoDB, Chatbot with RAG.
Developed a full-stack e-commerce web application using Vue.js, Node.js (Express.js), SQL, and PrimeVue UI Library. Led version control using GitHub, coordinated team collaboration, and deployed the project to the cloud via Docker on Render (PaaS).
I learned basic understanding of how full-stack developement works, system design patterns, development practice, responsive design, and mobile UI development.
I took the lead in selecting the right tech stack, designing the system architecture, and delegating tasks to each team member. However, driven by enthusiasm, I ended up handling most of the project myself, leaving only some CRUD backend tasks and the main shopping page to my teammates.
This project, “Dummy Chef Assistant: Generative AI-Controlled Recipe Recommendation System,” integrates Gemini API to detect image ingredient and algorithms to recommend suitable recipes based on the identified ingredients. By utilizing a precise algorithm to match ingredients with recipes, the system aims to provide recommendations that best meet user needs, enhancing user satisfaction. The system includes both a mobile app and a web interface, offering diverse and user-friendly operation platforms.
Built an AI-based ingredient recognition system that recommends suitable recipes. Developed the frontend interface and trained a model with Yolov8 to identify and analyze over 100 ingredients. Eventually, implemented the final version using Gemini for ingredient recognition and recipe recommendations.
For graduation topic paper and poster, please refer to my project portfolio drive below.
Developed an IoT system integrating NVIDIA Jetson Nano, Yolo, and ROS for real-world applications.
We have two bot, alpha and beta. Alpha will be controlled manually through controller and stops when it detects an obstacle. While beta will follow wherever alpha go through computer vision. When alpha wants to move backwards, it will send signal and beta will move backwards together to avoid collision between two bot.
Lead the development of the system, optimize performance of the system, and trained tiny Yolov4 model for image recognition.
Implement a Minecraft Launcher Desktop App with Tkinter and Python. The app allows users to launch Minecraft games and manage their settings.
The project is a group project. My friend BloodnightTW developed the Minecraft Launcher logic. While I developed the desktop app GUI, settings and skin selection functionality.
Made a simple shopping website with PHP and MySQL.
Made simple tetris-like game through Arduino and LED display on Practicum in Micropreprocessor course.
Implement logic gate design knowledge on a 7-seg display from binary and decode to display signals in my favourite game Minecraft.